Abstract: Can neuroscience dramatically improve our ability to design health communications? Modifiable health behaviors including poor diet, physical inactivity, and tobacco and alcohol consumption are leading causes of morbidity and mortaiity, both in the United Statesl and throughout the developed world2; yet changing these behaviors has proved an immensely challenging problem. Classic behavior change theories provide a foundation to develop and understand effective health campaigns and interventions;3 however there is still considerable variability in the effectiveness of such campaigns that we are unable to predict and explain. By improving our ability to understand and predict behavior change, neuroimaging methods such as functional magnetic resonance imaging (fMRl) may aid in the creation of maximally effective health campaigns. There may be important precursors of behavior change that are not easily obtained through self-reports, but that can be assessed with fMRl. In particular, people are notoriously limited in their ability to predict their own future o. behavior and accurately identiy their internal mental processes through verbal and written self-report Our team has found that activity in a prioridefined neural regions of interest can double the proportion of variance explained in individual behavior change following persuasive messaging, beyond self-report measures (e.9. attitudes, intentions, self-efficacy).5'6 The current proposal posits a next leap: neuroimaging technology may also be applied to more accurately forecast population level responses to health communications, and could dramatically improve the way that we design and select health communications. To this end, we propose to: (1) identify the neurocognitive signatures of health communications that are successful at changing behavior at the population level; (2) use these maps to forecast the success of new health messages; and, (3) use the information gained about underlying mechanisms of message success to advance theory and to develop novel strategies for message design. We will employ sophisticated multivariate and machine learning data analysis techniques (e.9. reinforcement learning models and pattern classification) to characterize the neural systems that are involved in processing successful health messages (i.e. messages that ultimately facilitate behavior change in larger, independent groups). Such techniques will provide insight about the mechanisms that lead messages to be optimally effective for populations on average, as well as helping to understand heterogeneity within populations (i.e. for whom are given messages likely to be most effective). These techniques will also allow us to define models that optimally combine neuroimaging data with other available data sources (e.9. self-report). Achievement of our goals (to identify neural patterns that predict message success and to test the psychological meaning of these activations) will facilitate the design and dissemination of more effective health messages, and will allow more efficient translation of core theoretical advances across behavior and disease specific silos. Public Health Relevance: Modifiable health behaviors including poor diet, physical inactivity, and tobacco and alcohol consumption are leading causes of morbidity and mortality, both in the United States1 and throughout the developed world2; yet changing these behaviors has proved an immensely challenging problem. The proposed program of research is designed to (1) identify the neurocognitive signatures of health communications that are successful at changing behavior at the population level; (2) use these maps to forecast the success of novel health messages; and, (3) use the information gained about underlying mechanisms that promote message success to advance theory. Achievement of our goals (to identify neural patterns that predict message success and to test the psychological meaning of these activations) will facilitate the design and dissemination of more effective health messages, and will allow more efficient translation of core theoretical advances across behavior and disease specific silos.